Abstract:
Shadow is an inseparable aspect of all natural scenes. When there are multiple light sources or multiple reflections several different shadows may overlap at the same location and create complicated patterns. Shadows are a potentially good source of information about a scene if the shadow regions can be properly identified and segmented. However, shadow region identification and segmentation is a difficult task and improperly identified shadows often interfere with machine vision tasks like object recognition and tracking. A shadow separation and contrast enhancement method based on the polarization of light is provided. Polarization information of scenes is captured by a polarization-sensitive camera and the scenes are processed to effectively separate shadows from different light sources.
Abstract:
An optical fingerprinting method extracts high quality latent fingerprints from a surface without any invasive chemical or physical contact with the examined object, and requires no cooperation of the subject. Rather than employing extraneous material, the optical properties of the latent fingerprint are used to generate one or more images with sufficient contrast to distinguish the latent fingerprint or some other deformation in the surface. The system includes a light source oriented to apply light at an angle of incidence to the surface at the position to be examined for the latent fingerprint or deformation, a camera oriented to receive light specularly and diffusely reflected from the surface and/or by the fingerprint or deformation on the surface, and a processor that performs the computation for digital contrast enhancement and/or reprojection of the recovered fingerprint image to a frontal view if necessary. The technique uses optical polarization properties to enhance the images by placing a linearly polarized filter(s) in front of the observing camera. At least two pictures of the same scene with the same lighting and view angle arrangement are taken whereby each of the pictures differ only in that the orientations of the polarization filter are different. At least two light polarization parameters for each pixel are computed from the two or more images taken with different polarizer orientations. An image with each pixel value representing the value of one of the polarization parameters or a function of the polarization parameters is generated and displayed, with some digital contrast enhancement and/or reprojection applied. The hidden latent fingerprint pattern is revealed in at least one such image with the interfering background pattern significantly suppressed.